Lucene search
+L

253 matches found

OSV
OSV
added 2026/06/29 11:50 a.m.12 views

PYSEC-2026-542 smolagents has Sandbox Escape Vulnerability in the local_python_executor.py Module

A sandbox escape vulnerability was identified in huggingface/smolagents version 1.14.0, allowing attackers to bypass the restricted execution environment and achieve remote code execution RCE. The vulnerability stems from the localpythonexecutor.py module, which inadequately restricts Python code...

9.9CVSS7.9AI score0.18654EPSS
SaveExploits1References6
RedhatCVE
RedhatCVE
added 2026/06/26 8:3 a.m.16 views

CVE-2026-41523

A flaw was found in vLLM, an inference and serving engine for large language models LLMs. An unauthenticated attacker can exploit an assert-based security check during activation function loading. By publishing a malicious HuggingFace model, an attacker can achieve arbitrary code execution on the...

7.5CVSS6.4AI score0.00746EPSS
SaveExploits1References6
Veracode
Veracode
added 2026/06/24 5:32 a.m.20 views

Out Of Band Data Exfiltration

Claude Code is vulnerable to Out-of-Band Data Exfiltration. The vulnerability is due to the pre-approval of the hostname huggingface.co as a bare hostname for the WebFetch tool, where any path on that domain—including attacker-controlled model repositories—was auto-approved without a permission...

9.1CVSS5.9AI score0.00403EPSS
SaveExploits0References1Affected Software1
OSV
OSV
added 2026/06/23 5:6 p.m.9 views

CVE-2026-54316 Claude Code: Out-of-Band Data Exfiltration via Pre-Approved HuggingFace Domain in WebFetch

Claude Code is an agentic coding tool. From 0.2.54 until 2.1.163, because the hostname huggingface.co was pre-approved as a bare hostname for the WebFetch tool, any path on that domain—including attacker-controlled model repositories—was auto-approved without a permission prompt or being subject ...

6CVSS6AI score
SaveExploits0References3
ATTACKERKB
ATTACKERKB
added 2026/06/23 5:6 p.m.11 views

CVE-2026-54316

Claude Code is an agentic coding tool. From 0.2.54 until 2.1.163, because the hostname huggingface.co was pre-approved as a bare hostname for the WebFetch tool, any path on that domain—including attacker-controlled model repositories—was auto-approved without a permission prompt or being subject ...

6CVSS5.9AI score0.00403EPSS
SaveExploits0References2Affected Software1
CVE
CVE
added 2026/06/23 5:6 p.m.115 views

CVE-2026-54316

CVE-2026-54316 (Claude Code) affects Claude Code versions 0.2.54–2.1.162, fixed in 2.1.163. The WebFetch tool allowed any path on the pre-approved domain huggingface.co, enabling an attacker-controlled repository path to be fetched without prompts or --allowedTools restrictions. If an attacker ca...

9.1CVSS5.9AI score0.00403EPSS
SaveExploits0References1Affected Software1
Cvelist
Cvelist
added 2026/06/23 5:6 p.m.43 views

CVE-2026-54316 Claude Code: Out-of-Band Data Exfiltration via Pre-Approved HuggingFace Domain in WebFetch

Claude Code is an agentic coding tool. From 0.2.54 until 2.1.163, because the hostname huggingface.co was pre-approved as a bare hostname for the WebFetch tool, any path on that domain—including attacker-controlled model repositories—was auto-approved without a permission prompt or being subject ...

6CVSS0.00403EPSS
SaveExploits0References1
NVD
NVD
added 2026/06/22 11:16 p.m.15 views

CVE-2026-41523

vLLM is an inference and serving engine for large language models LLMs. Prior to 0.22.0, an assert-based security check in vLLM's activation function loading allows any unauthenticated attacker to achieve arbitrary code execution on the server by publishing a malicious HuggingFace model, when vLL...

7.5CVSS0.00746EPSS
SaveExploits1References8
Vulnrichment
Vulnrichment
added 2026/06/22 10:18 p.m.15 views

CVE-2026-41523 vLLM: Security Check Bypass via assert Statement in Activation Function Loading Allows Arbitrary Code Execution

vLLM is an inference and serving engine for large language models LLMs. Prior to 0.22.0, an assert-based security check in vLLM's activation function loading allows any unauthenticated attacker to achieve arbitrary code execution on the server by publishing a malicious HuggingFace model, when vLL...

7.5CVSS6.5AI score0.00746EPSS
SaveExploits1References3
ATTACKERKB
ATTACKERKB
added 2026/06/22 10:18 p.m.9 views

CVE-2026-41523

vLLM is an inference and serving engine for large language models LLMs. Prior to 0.22.0, an assert-based security check in vLLM's activation function loading allows any unauthenticated attacker to achieve arbitrary code execution on the server by publishing a malicious HuggingFace model, when vLL...

7.5CVSS6.5AI score0.00746EPSS
SaveExploits1References4Affected Software1
OSV
OSV
added 2026/06/22 10:18 p.m.9 views

CVE-2026-41523 vLLM: Security Check Bypass via assert Statement in Activation Function Loading Allows Arbitrary Code Execution

vLLM is an inference and serving engine for large language models LLMs. Prior to 0.22.0, an assert-based security check in vLLM's activation function loading allows any unauthenticated attacker to achieve arbitrary code execution on the server by publishing a malicious HuggingFace model, when vLL...

7.5CVSS6.6AI score
SaveExploits0References10
OSSF Malicious Packages
OSSF Malicious Packages
added 2026/06/22 9:53 p.m.21 views

Malicious code in zod-pino (npm)

--- -= Per source details. Do not edit below this line.=- Source: amazon-inspector c536e5a7ee3d5542e1ac822b30ba4525e52b2ae0c964d0c2470468d91b9b41c8 The package is published under a name suggesting a Pino logger integration for Zod, but the tarball contents do not match that purpose and exhibit...

5.9AI score
SaveExploits0References7
OSV
OSV
added 2026/06/22 9:53 p.m.20 views

MAL-2026-6273 Malicious code in zod-pino (npm)

--- -= Per source details. Do not edit below this line.=- Source: amazon-inspector c536e5a7ee3d5542e1ac822b30ba4525e52b2ae0c964d0c2470468d91b9b41c8 The package is published under a name suggesting a Pino logger integration for Zod, but the tarball contents do not match that purpose and exhibit...

5.9AI score
SaveExploits0References7
OSSF Malicious Packages
OSSF Malicious Packages
added 2026/06/22 12:0 p.m.16 views

Malicious code in forge-jsx4 (npm)

forge-jsx4 is a remote access trojan RAT published to the public npm registry by the account rafaelsilva [email protected]. Versions 1.0.122 and 1.0.123 were published on June 21-22, 2026 as a reconstitution of the forge-jsx / forge-jsxy campaign, reusing the same command-and-control...

5.9AI score
SaveExploits0References3
OSV
OSV
added 2026/06/22 12:0 p.m.17 views

MAL-2026-6279 Malicious code in forge-jsx4 (npm)

forge-jsx4 is a remote access trojan RAT published to the public npm registry by the account rafaelsilva [email protected]. Versions 1.0.122 and 1.0.123 were published on June 21-22, 2026 as a reconstitution of the forge-jsx / forge-jsxy campaign, reusing the same command-and-control...

6.1AI score
SaveExploits0References3
Github Security Blog
Github Security Blog
added 2026/06/17 6:6 p.m.23 views

Claude Code: Out-of-Band Data Exfiltration via Pre-Approved HuggingFace Domain in WebFetch

Because the hostname huggingface.co was pre-approved as a bare hostname for the WebFetch tool, any path on that domain—including attacker-controlled model repositories—was auto-approved without a permission prompt or being subject to --allowedTools restrictions. An attacker able to inject untrust...

9.1CVSS5.5AI score0.00403EPSS
SaveExploits0References3Affected Software1
OSV
OSV
added 2026/06/17 6:6 p.m.19 views

GHSA-FG94-H982-F3MM Claude Code: Out-of-Band Data Exfiltration via Pre-Approved HuggingFace Domain in WebFetch

Because the hostname huggingface.co was pre-approved as a bare hostname for the WebFetch tool, any path on that domain—including attacker-controlled model repositories—was auto-approved without a permission prompt or being subject to --allowedTools restrictions. An attacker able to inject untrust...

9.1CVSS5.6AI score0.00403EPSS
SaveExploits0References3
Positive Technologies
Positive Technologies
added 2026/06/17 12:0 a.m.24 views

PT-2026-50594

Name of the Vulnerable Software and Affected Versions Claude Code versions 0.2.54 through 2.1.162 Description The WebFetch tool pre-approved the hostname 'huggingface.co' as a bare hostname, allowing any path on that domain to be auto-approved without a permission prompt or restrictions from...

9.1CVSS5.9AI score0.00403EPSS
SaveExploits0References7
Cvelist
Cvelist
added 2026/06/16 11:35 p.m.30 views

CVE-2026-48797 Backpropagate: backprop ui --auth and backprop ui --share do not enforce authentication

Backpropagate is a Python library for fine-tuning large language models on a single GPU. In versions 1.1.0 and 1.1.1, the optional Reflex web UI exposes a training control plane without authentication: dataset upload, model load, training start/stop, multi-run orchestration, GGUF export, and...

9.3CVSS0.00324EPSS
SaveExploits0References2
Github Security Blog
Github Security Blog
added 2026/06/16 5:34 p.m.78 views

vLLM: Security Check Bypass via assert Statement in Activation Function Loading Allows Arbitrary Code Execution

Summary An assert-based security check in vLLM's activation function loading allows any unauthenticated attacker to achieve arbitrary code execution on the server by publishing a malicious HuggingFace model, when vLLM runs in Python optimized mode python -O or PYTHONOPTIMIZE=1. Details vLLM uses ...

7.5CVSS7.7AI score0.00746EPSS
SaveExploits1References11Affected Software1
Rows per page
Query Builder